Existing liquid sensing solutions typically fail to achieve real-world scalability. Current channel-based methods (e.g., RF, acoustics) approach this problem through a consumer-centric classification paradigm. However, their measurements become tightly coupled to non-repeatable channel responses and heterogeneous hardware. As a result, systems require massive, ad-hoc data collection to cover these variations, yet still fail to generalize across the practically infinite diversity of products. In this paper, we propose a paradigm shift to targeted verification and repurpose the identity tag as a sensing interface. We present AquaCode, the first liquid quality assessment system seamlessly integrated into retail checkout platforms. Instead of inferring liquid types via unstable channels, AquaCode harnesses Liquid-Solid Interface (LSI), a novel sensing modality that measures charge transfer dynamics directly at the outer container wall, making it immune to external multipath interference. Crucially, AquaCode leverages the checkout platform to provide standardized mechanical excitation, converting the erratic motion artifacts of handheld sensing into a deterministic signal source. This enables a robust verification model: the system compares the runtime LSI signal against a manufacturer-defined golden standard encoded in a low-cost passive tag. We design a specialized analog frontend to extract these faint, high-impedance signals and a lightweight algorithm for 1-to-1 verification. We achieve a recall of 0.905 for identifying low-quality drinks. We believe AquaCode offers a scalable path to ubiquitous liquid safety.